PanchangaClockOnDiabetesAs of
A public board for type 2 diabetes

The Type 2 Board

Where the fight against type 2 diabetes stands, read as a game. The whole population is a game of Go: where is the open ground, and which groups are about to die? Single fights, like the search for a cure, get a chess close-up. The board shows what works, what doesn't yet, and where pooled effort would count most.

Observational: it records what real players doEvery claim linked to a sourceForecasts scored in public
1 · What should the next move be?

Hold remission past the first year

Many people can reach remission. Most lose it: in the DiRECT trial, 46% were in remission at 1 year and 13% at 5. That leak is the biggest open ground on the board, and it can be tested within a few years. Pooled effort goes there first. (University of Glasgow, Newcastle University)

  • Managed on drugs. Alive, as long as treatment continues.
  • In remission. Alive but weak, pressed by relapse.
  • ◎Big point. Holding remission past year 1.
  • △Vital point. Subtyping: who can remit, who can't.
  • Unclaimed ground. People not yet diagnosed.
Big pointFirst

Keep remission alive past year 1

After weight loss, the body defends its old weight: appetite rises and energy use falls. That is the likely main driver of the drop from 46% to 13%. People are often left alone right when support matters most.

  • ClinicsBuild follow-up and weight-maintenance support into every remission program.
  • PayersCover maintenance, not just the first 12 weeks.
  • ResearchersTest a GLP-1 drug as a bridge after remission against diet support alone.
  • FundersPay for 2- and 5-year follow-ups. RESET for Remission stopped at 24 weeks.
  • PatientsPeer groups can share what has helped people stay in remission.

Signals to watch: GLP-1 maintenance beats diet support at 2 years (B2, 45% by 2029). 30%+ stay in remission a year after stopping a GLP-1 drug (P3, 25% by 2029).

Vital pointSecond

Know which type 2 each person has

People differ in insulin resistance and in how much insulin-cell reserve is left. A test that tells who can remit, and who needs cell-level treatment, would aim every other move. Today it rarely changes care. (Chile study, 2026)

  • ResearchersRun trials that assign treatment by insulin-resistance and reserve scores, not weight alone.
  • FundersBack those trials. None with results was found as of Nov 2026.
  • CliniciansRecord reserve measures such as C-peptide, so the data exists to learn from.
  • PatientsJoin registries and trials that collect this data.

Signal to watch: a randomized trial assigning treatment by subtype reports results (P1, 30% by 2028).

Watch squareThird

The base: energy surplus and the food environment

GLP-1 drugs reach it by cutting appetite, but only while they're taken. Nothing reaches it at the scale of the food environment. That work is mostly policy and runs on a decade-long clock, so it is watched rather than pushed first.

  • Public healthTest changes to what's cheap and easy to eat, and measure the effect on type 2.
  • ResearchersStudy why fat comes back and why each person's fat limit differs.

The order follows the biggest measurable leak, not the biggest prize. "Who could help" lists are the board's reading of where effort would count, not endorsements. The board does not ask for or handle money.

How you can help

Every reader can move a stone

Each step below moves one of the squares above. The first one costs nothing and starts with one conversation.

Patients

Families and peers

  • Don't leave someone alone after remission. Check-ins, a walking buddy or a peer group go straight at the leak from 46% to 13%.Big point

Clinicians

  • Offer supervised remission programs to eligible newly diagnosed patients, like England's NHS program.Big point
  • Plan follow-up support after remission is reached, not only before.Big point

Researchers

  • Run trials on holding remission past year 1, with 2- and 5-year follow-up.Big point
  • Test treatment assigned by subtype or by insulin-resistance and reserve scores.Vital point

Funders and payers

  • Cover supervised remission programs and their maintenance phase. A program that could end the need for drugs is worth paying for, not just the drugs.Big point

Not medical advice. Talk with your doctor before changing diet, medicines or treatment. Remission programs are supervised for a reason.

2 · Where things stand

White is holding the position but can't attack the cause yet

Today's drugs control blood sugar while they're taken. GLP-1 drugs also cut appetite. Losing fat from the liver and pancreas puts some people into remission. No treatment in human trials targets the cause at its source, so checkmate is still out of sight. (T2D Network, 2026 review)

Chess close-up · the fight for a cure

The population is a Go game; this zooms in on one fight: White's pieces against the cause, drawn as Black's pawn chain.

Checkmate ~1%

A metabolism restored to healthy function, lasting and off all drugs, for most people, this decade.

Board forecast CM1
Functional win Normal

No complications and normal metabolic markers, even if still on a drug. Drug status is recorded separately.

Board forecasts FW1, FW2
Draw Hold

Managed on one easy pill, with no complications. Where most care aims today.

Board rule
Loss Harm

Progression to complications: heart, kidney, eyes, nerves.

Oxford, UKPDS

Remission: how often people get there

Routine care, ever (Hong Kong)6%
England's program, 1 year27%
DiRECT trial, 1 year46%
Adults 18 to 45, 24 weeks54%
DiRECT, still at 5 years13%

Sources: Hong Kong cohort · NHS program · DiRECT · RESET for Remission. Remission means normal blood sugar off drugs. It is not checkmate: the cause is still there, and about two-thirds relapse within about 3 years in routine care.

What drives type 2? Three open questions

Each is a separate yes/no forecast, because the answers can overlap: fat can drive most cases while a cell defect drives one subtype.

Liver and pancreas fat, above a personal limit, is the main driver for most people55%
An insulin-cell defect drives a subtype of 10%+ of people50%
Something else (gut, brain) drives a distinct subtype15%
Body fat explains more of the variation in risk than genes do70%

Board forecasts CF4, CF5, CF6 (check Dec 31, 2031) and TE1 (by 2030). Genes decide whose limit breaks; the environment decides whether it gets crossed. Evidence for the fat reading: DiRECT, ReTUNE.

Hidden forces: cards still face down

Black may hold cards nobody can see yet. Each stays face down until solid primary evidence turns it over. These are starting reads, not yet checked against primary sources.

  • Sleep loss and shift work. Moderate: cohort links, and short experiments where less sleep lowers insulin sensitivity.
  • Metabolism-disrupting chemicals. Associations in cohorts; causal proof in people is weak.
  • Chronic stress. Weak to moderate: cohort links and a plausible hormone path.
  • The gut microbiome. Early: associations and small, short-lived transplant studies.
  • Industry influence on food policy. Documented in past episodes; today's effect unmeasured.

A card turns face up when a trial shows that changing it changes type 2 risk or remission.

What changed recently

Each move is something a real player did. Black's best reply is what reality answers back.

Remission is strongest in young adults. A 12-week low-calorie diet plus supervised exercise put 54% of adults aged 18 to 45 into remission, against 4% on usual care. (News-Medical, Medical Daily)

Black's reply: relapse. The trial stopped at 24 weeks, so nobody knows yet how long it lasts.

Subtypes look like a spectrum. In Chile, the insulin-deficient subtype had more kidney damage, but added little to death risk beyond ordinary blood sugar control. Type 2 looks more like one spectrum of two failures than several separate diseases. (Diabetology & Metabolic Syndrome)

Black's reply: a better map doesn't change treatment until a drug targets one of the two failures.

Cheaper drugs abroad. Generic semaglutide launched in India in March, with prices up to 70% lower, and in Canada in May. China has none yet; US protection runs to about 2032. (Life Science Daily)

Black's reply: in the US, price and coverage stay the limit.

A stronger queen. Retatrutide, a weekly triple-hormone shot, cut A1c by up to 2 points and weight by 16.8% over 40 weeks in early type 2. (Diabetes UK)

Black's reply: it works while taken. It is control, not cure.

Worn-out insulin cells can partly recover. In islets from 21 donors with type 2, 13 regained their glucose response after three days at normal blood sugar. (Suleiman et al., Science Advances)

Black's reply: this is lab work. No regrowth or protection drug has reached an efficacy trial in people with diabetes.

3 · Honest scorecard

Every forecast is logged, dated and scored, misses included

Each forecast gets a probability, a check date and a rule for what counts as yes, confirmed before anything is scored. On the check date it is scored with a Brier score: 0 is perfect, 0.25 is a coin flip. Nothing has come due yet. Five short-dated forecasts start the scoring in early 2027. Forecasts that were replaced stay on the record, unscored.

0Forecasts logged
0Scored so far (2 replaced)
—Running Brier score
Jan 31, 2027First check date (S2)
#ForecastOddsCheck dateStatus

Our own misses so far

When we re-checked the board against current sources and an outside review, some things were wrong or out of date. They are shown here instead of quietly fixed.

  • Wrong The board first said nothing reaches the base of the chain. GLP-1 drugs cut appetite, so they do reach the energy surplus, while taken. Corrected after an outside review.Board correction, Nov 20, 2026
  • Wrong Canada's semaglutide patent did not lapse in January 2026. It lapsed years earlier over an unpaid fee; generics could be sold from January 2026 and launched May 14.The Logic
  • Stale Orforglipron (Foundayo) cash price tops out at $299 a month, not $349.LillyDirect
  • Narrower "Patent expired in India and China" holds for India only. China has no approved generic yet.Life Science Daily
  • Narrower "Two-thirds relapse in about 3 years" comes from a routine-care cohort where only 6% reached remission at all.2 Minute Medicine
  • Narrower "46.5% stop GLP-1 drugs within a year" applies to people with type 2 inside a weight-loss cohort.Consultant360

Anomalies: things that don't fit

Real results that don't fit the board are logged, not explained away.

  • In DiRECT, people who stayed diabetic lost as much liver and pancreas fat as those in remission. Their insulin cells had to restart too.Newcastle University, 2018
  • In ReTUNE, 8 of 12 people of normal weight went into remission after losing weight. The fat limit looks personal, not set by BMI.EurekAlert, 2021
  • ReTUNE weight loss is reported as about 12% in one source and about 6% in another. The published paper should settle it.EurekAlert 2021 · HCPLive 2022
  • Unconfirmed Many people get type 2 without much excess fat, notably lean patients in South and East Asia. This is the weakest case for the fat-limit reading. Not yet checked against a primary source.Raised in an outside review of this board, Nov 2026
  • After gastric bypass, more than 10 years of diabetes did not significantly lower remission. Remaining insulin production predicted it; years did not.de Cleva et al., 2021
4 · How to read the board

Three games, each at its own scale

The board is a way of seeing a hard problem whole: who is moving, what reality answers, and how close anyone is to winning. Go shows the whole population, chess zooms in on single fights, and poker tracks forces nobody can see yet. Nobody here makes moves. The board records them.

Go for the population

Go has no single target, only ground to hold. It shows where the open ground is (holding remission), which groups are about to die (remissions without support), and who nobody has reached (people not yet diagnosed).

Chess for single fights

Chess shows one target and the forcing lines toward it, like the search for a cure. It is a close-up, never a second full game at the same scale.

Poker for hidden forces

Black may hold cards nobody can see, like sleep loss or chemicals. Each stays face down until solid evidence turns it over.

White is everyone working on it

Researchers, drug makers, regulators, clinics, payers and patients. White's moves are what they actually do, logged with a source.

Black is reality's best reply

Relapse, cost, worn-out cells, the food environment. For every move, the board asks what the strongest answer would be.

Four outcomes

Checkmate is a metabolism restored to healthy function, lasting and off all drugs, measured by how the body works, not by weight. A functional win is no complications and normal markers, with drug status recorded separately. A draw is good management with no complications. A loss is complications.

The cause is a pawn chain, and a loop

Head: failing insulin cells and insulin resistance. Middle: liver and pancreas fat. Base: energy surplus, food environment and genes. Chains are beaten at the base. Real causes also feed back on each other, which is why a one-time fix fades.

Pawns wait on a watch list

A possible new piece stays off the board until a real event, such as a trial result, brings it on. Results that don't fit go in the anomaly log.

Open questions carry odds

Where the evidence hasn't settled, such as what drives type 2, the board gives each reading its own odds until evidence decides. A question stays only if its answer would change a forecast.

5 · Support this board

Keep the board running

This board is independent and free to read. Support keeps it running, updated and scored every day.

What your support pays for:

  • The daily check of new trials, approvals and papers against primary sources
  • Keeping the forecasts scored in public, misses included
  • Hosting, and keeping the page free for everyone

It does not fund medical research or a cure, and it buys no influence over what the board says. Sources and odds stay the same whoever gives. ClockOnDiabetes means the board keeps the clock on diabetes: how close the world is to restored health, scored every day.

Buy me a coffee

Not medical advice. Supporting the board doesn't change that, and the board can't answer questions about anyone's care.